Utilizing Modern Computer Development Tools
In Implementing The
Resource Consumption Model for Process Design (RCM)
Richard Jerz
St. Ambrose University1
Davenport, Iowa 52803
Abstract
Industrial engineers need to understand and use modern computational tools in advanced
research. The Resource Cost Model for Process Design (RCM) is the result of several years of
research into better models for analyzing and selecting process design alternatives. RCM
methodology was developed using a computer-based model that included several important
programming technologies: object-based and object-oriented techniques, a relational database,
and a third-party custom component. A comprehensive modeling of process alternatives would
have been very difficult without these technologies.
This paper focuses on the important computer-based technologies used to develop RCM.
Research objectives, an RCM overview, an explanation of technologies, and conclusions are
provided.
Keywords
Process design, cost analysis, object-oriented, SQL, relational databases, object linking.
1. Introduction and Research Objective
Several varieties of production economic models (e.g., return on investment analysis, break-even
analysis, cost estimating, and design for manufacture) aid production process design for product
manufacture [1]. These models, however, fail to integrate sufficiently the concepts of cost,
production cycle time, production capacity, and utilization. The methodologies typically rely
upon these factors being separately analyzed, but do not guarantee that they are. Some
methodologies use a narrow production volume range, or worse, one production volume in their
calculations, which limits additional insight into economies of scale.
1 This research was completed as a student writing my doctoral dissertation in Industrial Engineering at
The University of Iowa, Iowa City, Iowa.
1
The research objective was to develop a better methodology for analyzing process design
alternatives. It was believed that the new methodology needed to incorporate not only cost
analysis, but time and capacity analysis as well since these metrics affect each other. But industry
has learned that the lowest cost alternative is not necessarily the best. Some companies may
choose a more expensive process alternative because it provides a better time advantage (i.e.,
faster response). Some may choose a process because it provides greater reserve capacity (e.g.,
when anticipation of higher future product demand exists). RCM concentrates on thoroughly
analyzing cost, time, and utilization factors, and clearly illustrating results. It does not attempt to
resolve these conflicting criteria. A better analysis and representation of results, however, should
lead to better process design decisions.
RCM was designed to perform detailed calculations across a production volume range. Typically,
when one chooses to develop a higher level of analysis, analysis time increases. It was believed
that a computer-based model would be needed to perform calculations. Concern over the
representation of numeric results led to using graphical results. This research provides both an
analytical and a computer model.
A brief overview of RCM is provided to better appreciate the application of the specific
computer-based technologies. Discussion of RCM’s analytical model is provided in [2]. This
paper focuses on several specific computer-model technologies used to develop RCM, since they
represent significant changes in the way modern research can be conducted.
2. Overview of the Resource Consumption Model
RCM is a decision-support methodology that provides greater understanding, fidelity and
sensitivity analysis to process design analysis than do other traditional techniques. RCM’s
foundational concept is that part production consumes resources that can be translated into cost,
time, and utilization metrics. RCM accounts for all resources, which can be equipment, labor,
energy, material, tooling, and other consumables used by process designs alternatives. It
characterizes resources generically and avoids the need for terms such as “fixed costs,” “variable
costs,” “overhead,” and so forth. Figure 1 shows most of the parameters used to describe
resources grouped by function in RCM.
Knowing details about individual resource consumption provides information on where
improvement opportunity exists. A resource consumption perspective is not new; this is the
perspective taken by activity-based costing proponents. However, RCM recognizes that
resources can by consumed by production or the passage of time (i.e., they may spoil or become
inert when not used fast enough.) Resource consumption is also affected by overall system
behavior.
For each resource, RCM performs quantity-based, time-based, and system-based calculations for
many production volume points (in RCM, one-hundred points are used) to determine whether a
resource’s life is controlled by its inherent productive capacity, by time, or by the system in which
it is used. The results are plotted, as shown in Figure 2. On this unit cost graph, the controlling
cost is the maximum of all individual lines. RCM can perform calculations and display results for
any resource within an alternative. The user can begin to understand better the effect that an
individual resource has within the alternative.
RCM provides the ability to graph up to six resources on one graph, as shown in Figure 3. This
representation of results provides a better understanding about an alternative's component costs at
2
various production volumes. Resource calculations can be accumulated to compare alternatives.
Results are shown in both tabular and graphical formats. Figure 4 shows how RCM illustrates a
cost comparison of alternatives.
Figure 1. Model parameters by function
Average Part Cost($) vs Production Volume
Proj= P1, Alt= A1, Selected Res= R1
0.010
Quantity Constrained
Time Constrained
System Constrained
0.009
0.008
Average Part Cost($)
0.007
0.006
0.005
0.004
0.003
0.002
0.001
0.000
50000
100000
150000
200000
Production Volume
250000
300000
350000
Figure 2. RCM - cost analysis for an individual resource
Figure 2, Figure 3, and Figure 4 provide cost results. Similar graphs are produced for time and
utilization results. RCM has many ways in which resources and alternatives can be manipulated
and displayed [2]. Only a few are illustrated in this paper.
3
Average Part Cost($) vs Production Volume
Proj= P3, Selected Alt= Single Torch
8
SRS
SPS
SWG
SWW
SLW
Sing
7
Average Part Cost($)
6
5
4
3
2
1
0
10000
20000
30000
40000
50000
60000
Production Volume
70000
80000
90000
100000
Figure 3. Alternative component costs
Average Part Cost($) vs Production Volume
Proj= Which printer should be purchased? , for Selected Alternatives
5.0
Purchase Cannon
Purchase HP
Purchase Epson
4.5
4.0
Average Part Cost($)
3.5
3.0
2.5
2.0
1.5
1.0
0.5
0.0
500
1000
1500
2000
2500
3000
Production Volume
3500
4000
4500
5000
Figure 4. RCM - cost comparison of alternatives
3. Computer Program Development
Today, engineers need to be able to explore many process design scenarios in a timely manner.
One of the greatest advantages of modeling RCM on a computer is that sensitivity analysis can be
quickly performed. RCM is a computationally extensive methodology. The computer model
4
provides quick and easy access to all parameter values and results. Without powerful personal
computers and software, the development and understanding of RCM would not have occurred.
3.1.
Object-Based Programming Application
The modern computing environment uses a graphical user interface (GUI) and a mouse or similar
pointing device. It makes little sense to go back to the days of text-based programming
environments, which were common over much of the last fifty years. RCM was developed on a
personal computer running the Windows95 operating system. This graphical environment is
characterized as object-based, meaning that the user directs a program's actions by clicking on
different objects in the display window. For example, Windows programs provide a command
button labeled “Exit” or “Quit” that needs simply to be pressed to exit the application. To create
a program’s functionality, the programmer must master creating and using various visual objects,
such as command buttons, pull-down menus, grids, and forms. This is called “object-based
programming,” but should not be confused with object-oriented programming, described later.
Many software development products were available for the development of RCM. After
considering the specific needs for RCM and the direction of programming tools, Microsoft's
Visual FoxPro® was selected. This product not only provided an object-based development
environment, but also object-oriented programming and a database – both to be described below.
Other products at the time this research commenced, such as Visual Basic® and Microsoft
Access®, did not offer all of these capabilities.
A challenge for RCM was to design the user interface so that it was easy finding and changing any
model parameter values, and easy getting to various results. The solution was a tabbed-page
approach called a “page frame,” which is analogous to having tabs within a notebook. By using
this approach, the programmer can group similar information and the program’s work space is
great. The user can quickly tab to a page of interest. RCM's page approach is shown in Figure 5.
This figure shows only the “Data” page, which is one of six pages. The Plotting page gathers
information on what is to be plotted; alternatives or resources; how it is to be plotted; the
production volume range; and cost, time, and utilization axes ranges. “Cost,” “Time,” and
“Utilization” pages display graphic results, and the “Summary” page provides tabulated results.
The Data page contains three grids, called grid objects. Three grids were used because RCM data
was organized in three relational database tables. Functionally, the user selects a problem to
investigate in the “Projects” grid by clicking on it with the mouse. When this happens, all
alternatives that the problem contains appear in the second grid. When an alternative in the
“Alternative” grid is selected with the mouse, all resources that compose the alternative appear in
the “Resources” grid. The three grids make it easy to “zoom in” and “zoom out” to different
detail levels. Parameter values can also be changed within these grids.
An example of the Cost page is shown in Figure 6. It becomes obvious from Figure 5 and Figure
6 that the graphics environment and the use of objects accomplishes the goal of making RCM
easy to use and informative.
3.2.
Object-Oriented Programming
A discussion about object-oriented programming (OOP) is not complete without a discussion of
three specific terms always associated with OOP: inheritance, encapsulation, and polymorphism.
5
Figure 5. Data page
Figure 6. Cost page
Inheritance allows one to create new objects that are similar to, but a little different from existing
objects. For example, the programmer may want to define new command buttons that are similar
to, yet a little different from the standard button. The programmer may want the button’s length
and width to be sized specifically for the application. By “sub-classing” the standard button, a
6
programmer’s modification to the button’s size automatically changes all instances of the button
throughout the application.
Encapsulation, simply defined, is the ability to add properties (i.e., object features, such as value
parameters) and methods (i.e., program code) to objects that become active only when a specific
object is used, or “instantiated.” This is similar to declaring variables “local” or “private” within a
subroutine, but encapsulation is better. Instead of having an application made up of module after
module of spaghetti code, one object simply sends a message to another object, referencing that
second object's properties or methods [3]. The programming burden to remember whether
variable names are local or private is removed. Both data and code can be encapsulated within an
object. For example, in RCM, a “form” method named RCMCalculations was created to do most
of RCM's calculations and analysis. Being encapsulated within the form, the method is always
available whenever the form is active. Different objects within the RCM application communicate
their values and value changes with RCMCalculations automatically. RCMCalculations creates an
array (known in database terminology as a cursor) whenever needed. The Cost, Time, Utilization,
and Summary pages, when selected by the user, use this array in different ways to create their
visual results. These result pages need not worry about the data structure of the array - they can
simply use the array data as needed. Passing of data values is greatly simplified with
encapsulation.
Polymorphism is often thought to be a more difficult OOP concept than inheritance and
encapsulation, but it is really quite simple. It means that different objects can have methods or
properties with the same method name [4]. In other words, the programmer does not need to
worry about coming up with unique names for methods encapsulated in different objects. For
example, code that performs calculations for one command button can be called “Calculate,”
while other code that performs different calculations for another object can also be called
“Calculate.” Each object knows about its own “Calculate” routine. Code is executed by calling
the method with commands similar to “ThisButton.Calculate” or “ThatButton.Calculate.”
When properly implemented, OOP concepts improve the programmer's efficiency. However, it is
a shift from the traditional programming paradigm. For the industrial engineer, the ability to use
OOP techniques translates into faster application development. For example, at one point in the
development of RCM, the Utilization page did not exist. When the need for this page emerged, it
was added to RCM in approximately fifteen minutes of programming. Without OOP, this task
might have taken several hours to accomplish.
In its research form, RCM implements only a small subset of object-oriented programming. The
development of RCM into a commercially viable product would take greater advantage of OOP.
3.3.
Database Modeling Application and SQL
RCM needed to capture cost, time, and capacity information for every resource. A generic
characterization of parameters was believed to be beneficial for more efficient use and
management of information. Eventually, it was determined that fifteen parameters could
adequately define any resource. Since many resources can compose an alternative, and many
alternatives can compose a problem, data that RCM must manage can become enormous. A
database approach to the problem relieves the burden of information management [5].
A database approach for engineering process evaluation is not common to most traditional
methodologies. Most engineering analyses deal with variables and equations, not databases.
7
However, as the quantity of data available to companies grows, there is a need to understand
databases and information management concepts better.
In 1970, E.F. Codd, then a member of the IBM Research Laboratory in San Jose, California,
published a classic paper [6]: “A Relational Model of Data for Large Shared Data Banks.” It laid
down a set of abstract principles for database management: the so-called relational model. The
many advantages of the relational approach are too well known to need repeating here. Another
particular aspect to relational database research was the development of a standard relational
language to define, access, and manage databases. This language eventually became known as
SQL, an abbreviation for “Structured Query Language.” SQL is now an American National
Standards Institute (ANSI) standard [7]. SQL has been implemented in many database products.
RCM uses both the relational database approach and SQL to efficiently organize and manage its
data.
The SQL SELECT command is used to retrieve information. The basic form of SELECT has
three clauses: SELECT <attribute>, FROM <table>, and WHERE <condition>. The modifiers
( <> ) to the clauses can become very complex, and it often takes much practice to master all its
features. Because grouping and aggregation are required in many database applications, SQL has
included COUNT, SUM, MAX, MIN, and AVG functions. An example of SQL might help
understand its benefits.
RCM needed to calculate the overall controlling cycle time and the minimum availability for
alternatives based upon the alternative resource’s use of time. RCM does consider that some
resources’ cycle times can overlap with others, but the determination of overall cycle time and
minimum availability is not as simple as one might think. In many traditional programming
languages, the programmer would do these calculations using FOR/NEXT loops. In a database, a
solution using the SQL SELECT statement becomes easier. The code associated with these
calculations is shown in Figure 7
Although the code appears long and “wordy,” anyone who understands the SQL SELECT
statement would see that the code is simple. Some wordiness in code is the result of descriptive
object names that enhance readability. The advantage to using SQL over FOR/NEXT loops is
speed – most databases are optimized to execute SQL commands.
The ability to manipulate database information efficiently is not commonly considered a skill that
industrial engineers should have. However, engineers who possess this skill can manage
information more efficiently and make better decisions.
3.4.
Object Linking and Embedding (OLE)
RCM needed to generate specific graphs to represent results. However, Visual FoxPro’s
graphing capability was very limited. Facing this problem, the author had to decide whether to
develop a special purpose graphing routine, or to purchase a commercially available routine.
Since one primary research goal is to complete the research in a timely fashion, developing one’s
own code was an unattractive alternative.
One of Microsoft’s strategies for its Windows operating system is the ability of discrete
applications to talk to each other. Their vehicle to implement this ability is Object Linking and
Embedding, or OLE. OLE is not one tool; rather, it consists of two similar, but different methods
to allow applications to talk to each other. The first is the use of OLE controls in applications.
8
The idea is to be able to extend an application with third-party objects. These third party objects,
called ActiveX controls, custom controls, or OCX’s can be placed in any application that supports
the OLE standard. For instance, a graphics control written to the OLE standard could be placed
in a Visual FoxPro form, or a Visual Basic form, or any other design tool that also conforms to
the OLE specification. The second is OLE automation, or the ability of one application to run
another application. RCM takes advantage of the first type of OLE function.
* Calculate the overall controlling cycle time accounting for
* all overlaps.
* (used for system constraint calculations)
* First, get the overall time for groups in series without overlap
SELECT MAX(nresprodtime*(1-nrespcntover)/nresprodpcs) as ControlTime;
FROM rcm!resources;
WHERE Resources.cprojid = lcCurrentProjectID;
AND Resources.caltid = lcCurrentAlternativeID;
GROUP BY Resources.ngroup;
into cursor lnControlTime
* Combine the controlling sequence time and the largest individual
* resource time for an alternative.
select sum(controlTime);
from lnControlTime;
union;
SELECT MAX(Resources.nresprodtime/nresprodpcs);
FROM rcm!resources;
WHERE Resources.cprojid = lcCurrentProjectID;
AND Resources.caltid = lcCurrentAlternativeID;
into cursor lnControlTime
* Select the largest time between the sequential resource times and
* the largest individual resource time.
* This become the controlling cycle time.
select max(sum_controltime) as CycleTime;
from lnControlTime;
into array lnControlTime
* Now get the minimum resource availability for unit costs (not batch).
SELECT MIN(Resources.nresavail) as MinAvail;
FROM rcm!resources;
WHERE Resources.cprojid = lcCurrentProjectID;
AND Resources.caltid = lcCurrentAlternativeID ;
AND Resources.lbatch = .F.;
INTO array lnMinAvailability
Note: the *’s in the code are only comment statements.
Figure 7. SQL SELECT code
For this research, a special graphing control produced by Gigasoft, Inc. called ProEssentials [8]
was purchased. The cost graph shown above in Figure 6 is a ProEssentials object. Functionally,
the graph control contains many properties that define its appearance. More important, data
generated by RCM's methods are simply communicated to the graph object. The program code
required to produce the cost lines in Figure 6 is shown below.
9
thisform.pgfMain.Page3.olegphCost.XData[k,
thisform.pgfMain.Page3.olegphCost.YData[0,
thisform.pgfMain.Page3.olegphCost.YData[1,
thisform.pgfMain.Page3.olegphCost.YData[2,
i-1]
i-1]
i-1]
i-1]
=
=
=
=
lnProdVolume
lnCost[1],0
lnCost[2],0)
lnCost[3],0)
Table 1 provides an example of some ProEssentials property settings as they occur within RCM
and a description of what they do.
Table 1. Graph object properties and function in RCM
ProEssentials Command
.points = 100
.graphbackcolor = RGB(0,0,0)
.subsetcolors(0) = RGB(255,255,0)
.subsetpointtypes(0) = 3
.subsetLineTypes(5) = 1
.xaxislabel = “Production Volume”
Function
Plot 100 points
Sets background to black
Sets color of the first plotted line
Sets first line point type to a solid circle
Sets sixth line type to a dashed line
X-axis label
Only a small subset of ProEssentials capabilities was needed for this research. ProEssentials can
also produce many other charts, including bar charts, logarithmic charts, pie charts, polar charts,
Pareto charts, and real-time data tracking charts. It is a product that serves the engineering
community well. It is important that industrial engineers be aware of third-party custom controls
and OLE so that one does not reinvent the wheel, and so that applications can be developed more
quickly.
4. Conclusions
The programming environment and the set of tools engineers have available today have changed
dramatically over the last fifty years. More complex research analysis and models can be
developed and their results communicated more effectively. A comprehensive modeling of
process alternatives for RCM would have been very difficult without these technologies.
The graphical user interface is here to stay. It has replaced writing lines of codes with an objectdevelopment environment. Any engineer who programs computer applications should master
graphical environment tools.
There are many advantages to understanding databases and the SQL standard. Engineers will
have greater access to and management of company information. Knowledge of SQL will allow
one to move from one database environment to another without the need for expensive retraining.
Applications written by engineers will have greater portability to different products, and the
lifetime of applications will be longer.
Engineers should be aware of third-party custom controls. Custom controls reduce programming
time and improve the quality and reliability of the application. In the future, RCM may implement
OLE automation. RCM may eventually communicate directly with other applications, such as
computer-aided design (CAD), resource planning and scheduling, financial planning, and other
computer-aided process planning systems.
10
5. Acknowledgments
This work has been partially supported by a predoctoral fellowship from the United States
Department of Energy (DOE) in “Integrated Manufacturing,” 1995.
References
[1]
Vonderembse, M.A. and White, G.P., Operations Management: Concepts, Methods, and Strategies,
3rd ed., West Publishing Company, New York, 1996.
[2]
Jerz, R., “A Resource Consumption Model (RCM) for Process Design,” Ph.D. Dissertation, The
University of Iowa, Iowa City, Iowa, December 1997.
[3]
Hentzen, W., Programming Visual FoxPro 3.0, Ziff-Davis Press, Emeryville, California, 1995.
[4]
Granor, T.E., and Roche, T., Hacker's Guide to Visual FoxPro 3.0, Addison-Wesley Developers
Press, New York, 1996.
[5]
Elmasri, R. and Navathe, S.B., Fundamentals of Database Systems, 2nd ed., Addison-Wesley
Publishing Company, Menlo Park, California, 1994.
[6]
Codd, E.F., “A Relational Model of Data for Large Shared Data Banks,” Communications of the
ACM, Vol. 13, No. 6, June 1970.
[7]
Date, C.J. and Darwen, H., A Guide to The SQL Standard, 3rd ed., Addison-Wesley Publishing
Company, New York, 1994.
[8]
Dede, R., GigaSoft ProEssentials User's Guide, Gigasoft, Inc., Keller, Texas, 1996.
11